Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Content-Type: text/plain;
        charset=utf-8
Mime-Version: 1.0 (1.0)
Subject: Re: [patch 09/11] x86/vdso: Simplify the invalid vclock case
From:   Andy Lutomirski <luto@amacapital.net>
In-Reply-To: <alpine.DEB.2.21.1809271630470.8118@nanos.tec.linutronix.de>
Date:   Thu, 27 Sep 2018 07:39:06 -0700
Cc:     John Stultz <john.stultz@linaro.org>,
        Andy Lutomirski <luto@kernel.org>,
        LKML <linux-kernel@vger.kernel.org>, X86 ML <x86@kernel.org>,
        Peter Zijlstra <peterz@infradead.org>,
        Matt Rickard <matt@softrans.com.au>,
        Stephen Boyd <sboyd@kernel.org>,
        Florian Weimer <fweimer@redhat.com>,
        "K. Y. Srinivasan" <kys@microsoft.com>,
        Vitaly Kuznetsov <vkuznets@redhat.com>,
        devel@linuxdriverproject.org,
        Linux Virtualization <virtualization@lists.linux-foundation.org>,
        Paolo Bonzini <pbonzini@redhat.com>,
        Arnd Bergmann <arnd@arndb.de>, Juergen Gross <jgross@suse.com>
Content-Transfer-Encoding: quoted-printable
Message-Id: <ABD3F880-C028-431B-A70D-E94D7F73154D@amacapital.net>
References: <20180914125006.349747096@linutronix.de> <20180914125118.909646643@linutronix.de> <CALCETrXh70oFHfT0Z-JXcnAumhAgLqSea5UiPY3WXu=v2GcKTg@mail.gmail.com> <CALAqxLUv9Rd_eA7u9gZNFvTNrT1Z67RpHLozdeNC0W2yZm=Heg@mail.gmail.com> <alpine.DEB.2.21.1809180948570.3558@nanos.tec.linutronix.de> <863331ED-B04A-4B94-91A2-D34002C9CCDC@amacapital.net> <alpine.DEB.2.21.1809190037570.1468@nanos.tec.linutronix.de> <439A3E73-E4FF-4D66-800E-5BEE58EDE8F6@amacapital.net> <alpine.DEB.2.21.1809190112020.1468@nanos.tec.linutronix.de> <alpine.DEB.2.21.1809271630470.8118@nanos.tec.linutronix.de>
To:     Thomas Gleixner <tglx@linutronix.de>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


> On Sep 27, 2018, at 7:36 AM, Thomas Gleixner <tglx@linutronix.de> wrote:
>=20
>> On Wed, 19 Sep 2018, Thomas Gleixner wrote:
>> On Tue, 18 Sep 2018, Andy Lutomirski wrote:
>>>> On Sep 18, 2018, at 3:46 PM, Thomas Gleixner <tglx@linutronix.de> wrote=
:
>>>>> On Tue, 18 Sep 2018, Andy Lutomirski wrote:
>>>>> Do we do better if we use signed arithmetic for the whole calculation?=

>>>>> Then a small backwards movement would result in a small backwards resu=
lt.
>>>>> Or we could offset everything so that we=E2=80=99d have to go back sev=
eral
>>>>> hundred ms before we cross zero.
>>>>=20
>>>> That would be probably the better solution as signed math would be
>>>> problematic when the resulting ns value becomes negative. As the delta i=
s
>>>> really small, otherwise the TSC sync check would have caught it, the ca=
ller
>>>> should never be able to observe time going backwards.
>>>>=20
>>>> I'll have a look into that. It needs some thought vs. the fractional pa=
rt
>>>> of the base time, but it should be not rocket science to get that
>>>> correct. Famous last words...
>>>>=20
>>>=20
>>> It=E2=80=99s also fiddly to tune. If you offset it too much, then the fa=
ncy
>>> divide-by-repeated-subtraction loop will hurt more than the comparison t=
o
>>> last.
>>=20
>> Not really. It's sufficient to offset it by at max. 1000 cycles or so. Th=
at
>> won't hurt the magic loop, but it will definitely cover that slight offse=
t
>> case.
>=20
> I got it working, but first of all the gain is close to 0.
>=20
> There is this other subtle issue that we've seen TSCs slowly drifting apar=
t
> which is caught by the TSC watchdog eventually, but if it exeeds the offse=
t
> _before_ the watchdog triggers, we're back to square one.
>=20
> So I rather stay on the safe side and just accept that we have to deal wit=
h
> that. Sigh.
>=20
>=20

Seems okay to me. Oh well.=20=